Optimization of the sample preparation and extraction methodology

Transcription

1 Supporting information Results and discussion Optimization of the sample preparation and extraction methodology Visual comparison of the slopes (curve in standard mixture vs MMC curve with identical concentration range) confirmed the presence of signal suppression due to matrix effects and necessitated further use of MMC curves for all three juice matrices and the different types of tomato products. Upon construction of MMC curves, significantly different slopes (nonparallelism of the curves, confirmed y t-test 1 ) were oserved for the majority of target analytes. This revealed the necessity to use matrix specific MMC curves for quantitation purposes (Figure 1). Method validation In Tales 1 and 2, regression coefficients (R 2 ) and experimental p-values from lack-of-fit tests for every analyte in each investigated matrix are summarized in Tales 1-2. Additionally, LOD and LOQ values are represented. Homoscedasticity was assessed as previously descried. 2 Briefly, homoscedasticity is evaluated y applying an F-test. If the experimental F-value is higher than the taled F-value, this is indicative of an heteroscedastic situation, which can e counteracted through weighted least squares linear regression (WLSLR). The optimal weighting factor, w i, is chosen according to a percentage relative error %RE: %RE=([C experimental -C assigned ]/C assigned )*100 The effectiveness of a weighting factor is evaluated y calculating %RE (the sum of asolute %RE values). In Tale 3, %RE and accuracy (in terms of ias, %) at three

2 concentration levels otained y using unweighted (w i = 1) and weighted (w i = 1/x 2 ) linear regression for all target analytes in tomato juice and tomato sauce are displayed. The weighting factor 1/x 2 not only produced the least %RE for these data sets, ut also consideraly improved the accuracy for the majority of analytes, particularly at the lowest concentration level of the caliration curve. Apparent recovery, RSD r, RSD R and U values for every analyte in each investigated matrix are displayed in Tales 4-5. Alternaria toxins in commercially availale foodstuffs This study reports the novel detection of modified Alternaria toxins (specifically, sulfates of (1) alternariol and (2) alternariol monomethyl ether) occurring in tomato products. Particularly in tomato concentrate, alternariol-3-sulfate (8) and alternariol monomethyl ether (10) were detected in 26% and 78% of all samples, in concentrations up to 8.7 and 9.9 µg/kg, respectively. A Synapt G2-Si mass spectrometer, operated in high resolution MS E continuum mode (ESI - ), was used to analyse tomato product samples from the survey in which sulfates of 1 and 2 were reported y low resolution tandem mass spectrometry. Accurate masses of oth sulfates with an acceptale mass deviation (< 2 mda) were detected in low energy as well as high energy mode. Component identification was performed y comparing the retention time under identical chromatographic conditions and y matching the high energy fragmentation spectra of the precursor ion from spiked samples to that of naturally contaminated samples (Figure 2). Chromatographic separation was performed using a Waters Acquity UPLC system (Waters, Milford, MA) equipped with a FTN autosampler. A sample volume of 5 μl was injected into an HSS T3 column (1.8 μm, 2.1 x 100 mm) held at 35

3 C with a flow rate of 400 μl/min. A gradient elution program with solvent A (ultra-pure water, 1% acetic acid) and solvent B (acetonitrile, 1% acetic acid) was applied as follows: 95% A and 5% B for 0.5 min followed y an increase to 95% B from 0.5 to 16.0 min, 95% B maintained from 16.0 to 17.0 min, ramping ack to 95% A from 17.0 to 17.1 min, and maintaining starting conditions from 17.1 to 20 min. Mass spectrometric detection was performed using a SYNAPT G2-Si (Waters, Milford, MA) equipped with an electrospray ionization source operating in negative mode with a capillary voltage of 2.5 kv and a sampling cone voltage of 30 V. The full-scan data were acquired in MS E continuum high resolution mode within a 50 to 1200 Da mass range with a 0.1 s survey scan time over a 17.5 min run time. In high energy mode, the trap MS collision energy was ramped from 30.0 to 50.0 ev. Desolvation temperature was 500 C, source temperature 150 C, cone gas flow 150 L/h and desolvation gas flow 1000 L/h. During acquisition, accurate masses were generated through correction using an external reference (Lock Spray, a 1 ng/μl solution of leucine encephalin infused at a flow rate of 10 μl/min) via a lock spray interface, generating a reference ion of m/z ([M H] ) in negative ionization mode.

4 References 1. Van Loco, J. Method Validation for Food Analysis: Concepts and Use of Statistical Techniques. In The Determination of Chemical Elements in Food: Applications for Atomic and Mass Spectrometry; Caroli, S., Eds.; John Wiley & Sons, Inc.: Hooken, NJ, USA, 2007; pp Walravens, J.; Mikula, H.; Asam, S.; Rychlik, M.; Njume Ediage, E.; Diana Di Mavungu, J.; Van Landschoot, A.; Vanhaecke, L.; De Saeger, S. Development and validation of an ultra-highperformance liquid chromatography tandem mass spectrometric method for the simultaneous determination of free and conjugated Alternaria toxins in cereal-ased foodstuffs. J. Chromatogr. A 2014, 1372,

5 Figure captions Figure 1. Non-parallelism (confirmed y t-test) 1 of the matrix matched caliration (MMC) curves of oth [A] 7 and [B] 4 in tomato juice versus tomato concentrate, [C] 1 in apple juice versus grape juice and [D] 8 in grape juice versus carrot juice. Parallelism 1 of the MMC curves of [E] 2 in tomato juice versus tomato concentrate and [F] 3 in tomato juice versus tomato paste due to the application of the corresponding isotope-laelled internal standards [ 2 H 4 ]-2 and [ 13 C 6, 15 N]-3. Figure 2. Extracted ion chromatogram of 10 (m/z ) in [A] spiked tomato product sample (MS E high energy mode), [B] spiked tomato product sample (MS E low energy mode), [C] naturally contaminated tomato concentrate sample (MS E high energy mode) and [D] naturally contaminated tomato concentrate sample (MS E low energy mode). Comparison of fragmentation spectra (MS E high energy mode) of 10 (m/z ) in [E] spiked tomato product sample and [F] naturally contaminated tomato concentrate sample.

6 Tales Tale 1. R² values and p-values (lack-of-fit test, SPSS) of the matrix-matched caliration curves (range µg/kg) in fruit and vegetale juices (apple, carrot and grape juice), supplemented with limits of detection (LOD) and limits of quantitation (LOQ) for all the analytes (µg/kg). Alternaria toxins a Apple juice Carrot juice Grape juice R² p LOD LOQ R² p LOD LOQ R² p LOD LOQ a 1: alternariol - 2: alternariol monomethyl ether - 3: tenuazonic acid - 4: tentoxin - 5: altenuene - 6: altertoxin-i - 7: alternariol-3- glucoside - 8: alternariol-3-sulfate - 9: alternariol monomethyl ether-3-glucoside - 10: alternariol monomethyl ether-3-sulfate.

7 Tale 2. R² values and p-values (lack-of-fit test, SPSS) of the matrix-matched caliration curves (range µg/kg) in lyophilised tomato products (juice, sauce and concentrate), supplemented with limits of detection (LOD) and limits of quantitation (LOQ) for all the analytes (µg/kg, expressed on fresh weight of the tomato products applying the experimentally determined conversion factor [CF]). Alternaria toxins a Tomato juice (CF=0.052) Tomato sauce (CF=0.077) Tomato concentrate (CF=0.216) R² p LOD LOQ R² p LOD LOQ R² p LOD LOQ a 1: alternariol - 2: alternariol monomethyl ether - 3: tenuazonic acid - 4: tentoxin - 5: altenuene - 6: altertoxin-i - 7: alternariol-3- glucoside - 8: alternariol-3-sulfate - 9: alternariol monomethyl ether-3-glucoside - 10: alternariol monomethyl ether-3-sulfate. Because of depletion of the stock solution of 6, validation experiments for ATX-I in tomato sauce and concentrate were not performed.

8 Tale 3. Sum of the relative errors ( %RE) and accuracy (Bias, %) at low (50 µg/kg), medium (250 µg/kg) and high (1000 µg/kg) concentration level otained y using unweighted (w i = 1) and weighted (w i = 1/x 2 ) linear regression for all the target analytes in tomato juice and tomato sauce. Alternaria toxins a w i %RE Tomato juice Tomato sauce Bias (%) Bias (%) w i %RE low medium high low medium high /x /x /x /x /x /x /x /x /x /x /x /x /x /x /x /x /x /x /x /x a 1: alternariol - 2: alternariol monomethyl ether - 3: tenuazonic acid - 4: tentoxin - 5: altenuene - 6: altertoxin-i - 7: alternariol-3- glucoside - 8: alternariol-3-sulfate - 9: alternariol monomethyl ether-3-glucoside - 10: alternariol monomethyl ether-3-sulfate. Because of depletion of the stock solution of 6, validation experiments for ATX-I in tomato sauce and concentrate were not performed.

9 Tale 4. Repeataility (RSD r ), intermediate precision (RSD R ), apparent recovery (R A, %) and expanded measurement uncertainty (U, %) values for all the analytes at low, medium and high concentration level (µg/kg) in fruit and vegetale juices (apple, carrot and grape juice). Type of juice Concentration (µg/kg) RSD r RSD R 7 a 8 a 5 a 9 a 3 a R A U RSD r RSD R R A U RSD r RSD R R A U RSD r RSD R R A U RSD r RSD R R A U Apple Carrot Grape Type of juice Concentration (µg/kg) RSD r RSD R 6 a 1 a 4 a 10 a 2 a R A U RSD r RSD R R A U RSD r RSD R R A U RSD r RSD R R A U RSD r RSD R R A U Apple Carrot Grape a 1: alternariol - 2: alternariol monomethyl ether - 3: tenuazonic acid - 4: tentoxin - 5: altenuene - 6: altertoxin-i - 7: alternariol-3-glucoside - 8: alternariol-3-sulfate - 9: alternariol monomethyl ether-3-glucoside - 10: alternariol monomethyl ether-3-sulfate. RSD r and RSD R acceptance criteria: 20 and 25%, respectively; R A imposed guideline ranges: %.

10 Tale 5. Repeataility (RSD r ), intermediate precision (RSD R ), apparent recovery (R A, %) and expanded measurement uncertainty (U, %) values for all the analytes at low, medium and high concentration level (µg/kg) in lyophilised tomato products (juice, sauce and concentrate). Tomato product Concentration (µg/kg) RSD r RSD R 7 a 8 a 5 a 9 a 3 a R A U RSD r RSD R R A U RSD r RSD R R A U RSD r RSD R R A U RSD r RSD R R A U Juice Sauce Concentrate Tomato product Concentration (µg/kg) RSD r RSD R 6 a,c 1 a 4 a 10 a 2 a R A U RSD r RSD R R A U RSD r RSD R R A U RSD r RSD R R A U RSD r RSD R R A U Juice Sauce Concentrate a 1: alternariol - 2: alternariol monomethyl ether - 3: tenuazonic acid - 4: tentoxin - 5: altenuene - 6: altertoxin-i - 7: alternariol-3-glucoside - 8: alternariol-3-sulfate - 9: alternariol monomethyl ether-3-glucoside - 10: alternariol monomethyl ether-3-sulfate. RSD r and RSD R acceptance criteria: 20 and 25%, respectively; R A imposed guideline ranges: %. c Because of depletion of the ATX-I stock solution, validation experiments for ATX-I in tomato sauce and concentrate were not performed.

11 Figure graphics Figure 1 Response Response Response Response Response

12 Figure 2 A B C D E F